RESEARCH ARTICLE Biogeography and environmental conditions shape bacteriophage-bacteria networks across the human microbiome Geoffrey D. Hannigan1, Melissa B. Duhaime2, Danai Koutra3, Patrick D. Schloss1* 1 Department of Microbiology and Immunology, University of Michigan, Ann Arbor, Michigan, United States of America, 2 Department of Ecology and Evolutionary Biology, University of Michigan, Ann Arbor, Michigan, United States of America, 3 Department of Computer Science, University of Michigan, Ann Arbor, Michigan, United States of America a1111111111 * [email protected] a1111111111 a1111111111 a1111111111 a1111111111 Abstract Viruses and bacteria are critical components of the human microbiome and play important roles in health and disease. Most previous work has relied on studying bacteria and viruses independently, thereby reducing them to two separate communities. Such approaches are OPEN ACCESS unable to capture how these microbial communities interact, such as through processes Citation: Hannigan GD, Duhaime MB, Koutra D, that maintain community robustness or allow phage-host populations to co-evolve. We Schloss PD (2018) Biogeography and implemented a network-based analytical approach to describe phage-bacteria network environmental conditions shape bacteriophage- diversity throughout the human body. We built these community networks using a machine bacteria networks across the human microbiome. PLoS Comput Biol 14(4): e1006099. https://doi. learning algorithm to predict which phages could infect which bacteria in a given micro- org/10.1371/journal.pcbi.1006099 biome. Our algorithm was applied to paired viral and bacterial metagenomic sequence sets Editor: Nicola Segata, University of Trento, ITALY from three previously published human cohorts. We organized the predicted interactions into networks that allowed us to evaluate phage-bacteria connectedness across the human Received: October 4, 2017 body. We observed evidence that gut and skin network structures were person-specific and Accepted: March 21, 2018 not conserved among cohabitating family members. High-fat diets appeared to be associ- Published: April 18, 2018 ated with less connected networks. Network structure differed between skin sites, with Copyright: © 2018 Hannigan et al. This is an open those exposed to the external environment being less connected and likely more suscepti- access article distributed under the terms of the ble to network degradation by microbial extinction events. This study quantified and con- Creative Commons Attribution License, which trasted the diversity of virome-microbiome networks across the human body and illustrated permits unrestricted use, distribution, and reproduction in any medium, provided the original how environmental factors may influence phage-bacteria interactive dynamics. This work author and source are credited. provides a baseline for future studies to better understand system perturbations, such as Data Availability Statement: All relevant data are disease states, through ecological networks. within the paper and its Supporting Information files. Funding: GDH was supported in part by the Molecular Mechanisms in Microbial Pathogenesis Author summary Training Program (T32AI007528). GDH and PDS were supported in part by funding from the NIH The human microbiome, the collection of microbial communities that colonize the (P30DK034933, U19AI09087, and U01AI124255). human body, is a crucial component to health and disease. Two major components of the The funders had no role in study design, data human microbiome are the bacterial and viral communities. These communities have pri- collection and analysis, decision to publish, or marily been studied separately using metrics of community composition and diversity. preparation of the manuscript. PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006099 April 18, 2018 1 / 24 Network diversity of the healthy human microbiome Competing interests: The authors have declared that no competing interests exist. These approaches have failed to capture the complex dynamics of interacting bacteria and phage communities, which frequently share genetic information and work together to maintain ecosystem homestatsis (e.g. kill-the-winner dynamics). Removal of bacteria or phage can disrupt or even collapse those ecosystems. Relationship-based network approaches allow us to capture this interaction information. Using this network-based approach with three independent human cohorts, we were able to present an initial understanding of how phage-bacteria networks differ throughout the human body, so as to provide a baseline for future studies of how and why microbiome networks differ in disease states. Introduction Viruses and bacteria are critical components of the human microbiome and play important roles in health and disease. Bacterial communities have been associated with disease states, including a range of skin conditions [1], acute and chronic wound healing conditions [2, 3], and gastrointestinal diseases, such as inflammatory bowel disease [4, 5], Clostridium difficile infections [6] and colorectal cancer [7, 8]. Altered human viromes (virus communities consist- ing primarily of bacteriophages) also have been associated with diseases and perturbations, including inflammatory bowel disease [5, 9], periodontal disease [10], spread of antibiotic resistance [11], and others [12±17]. Viruses act in concert with their microbial hosts as a single ecological community [18]. Viruses influence their living microbial host communities through processes including lysis, host gene expression modulation [19], influence on evolutionary processes such as horizontal gene transfer [20] or antagonistic co-evolution [21], and alteration of ecosystem processes and elemental stoichiometry [22]. Previous human microbiome work has focused on bacterial and viral communities, but have reduced them to two separate communities by studying them independently [5, 9, 10, 12±17]. This approach fails to capture the complex dynamics of interacting bacteria and phage communities, which frequently share genetic information and work together to maintain eco- system structure (e.g. kill-the-winner dynamics that prevent domination by a single bacte- rium). Removal of bacteria or phages can disrupt or even collapse those ecosystems [18, 23± 32]. Integrating these datasets as relationship-based networks allow us to capture this complex interaction information. Studying such bacteria-phage interactions through community-wide networks built from inferred relationships begins to provide us with insights into the drivers of human microbiome diversity across body sites and enable the study of human microbiome network dynamics overall. In this study, we characterized human-associated bacterial and phage communities by their inferred relationships using three published paired virus and bacteria-dominated whole com- munity metagenomic datasets [13, 14, 33, 34]. We leveraged machine learning and graph the- ory techniques to establish and explore the human bacteria-phage network diversity therein. This approach built upon previous large-scale phage-bacteria network analyses by inferring interactions from metagenomic datasets, rather than culture-dependent data [28], which is limited in the scale of possible experiments and analyses. We implemented an adapted metage- nomic interaction inference model that made some improvements upon previous phage-host interaction prediction models. Previous approaches have utilized a variety of techniques, such as linear models that were used to predict bacteria-phage co-occurrence using taxonomic assignments [35], and nucleotide similarity models that were applied to both whole virus genomes [36] and clusters of whole and partial virus genomes [37]. Our approach uniquely PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006099 April 18, 2018 2 / 24 Network diversity of the healthy human microbiome included protein interaction data and was validated based on experimentally determined posi- tive and negative interactions (i.e. who does and does not infect whom). We built on previous modeling work as a means to our ends, and focused on the biological insights we could gain instead of building a superior model and presenting our work as a toolkit. We therefore did not focus on extensive benchmarking against other existing models [36, 37±40]. Through this approach we were able to provide an initial basic understanding of the network dynamics asso- ciated with phage and bacterial communities on and in the human body. By building and uti- lizing a microbiome network, we found that different people, body sites, and anatomical locations not only support distinct microbiome membership and diversity [13, 14, 33, 34, 41± 43], but also support ecological communities with distinct communication structures and robustness to network degradation by extinction events. Through an improved understanding of network structures across the human body, we aim to empower future studies to investigate how these communities dynamics are influenced by disease states and the overall impact they may have on human health. Results Cohort curation and sample processing We studied the differences in virus-bacteria interaction networks across healthy human bodies by leveraging previously published shotgun sequence datasets of purified viral metagenomes (viromes) paired with bacteria-dominated whole community metagenomes. Our study con- tained three datasets that explored the impact of diet on